Four areas of research are proposed that systematically evaluate age-related differences in associative learning within a metacognitive framework. A central assumption of this framework is that people actively participate in learning new material by (1) utilizing their knowledge about cognitively demanding tasks to select rehearsal strategies, (2) monitoring on-going learning and performance, and (3) utilizing such monitoring to regulate control processes that govern learning and performance. Our long-term objectives include discovering how each of these aspects of self-directed learning influences the rate of learning and age differences in learning. The proposed research involves four separate areas of effort. In general, the proposed experiments involve the use of paired-associate learning and recall to study metacognition, strategies, and learning. Area 1 represents a critical step toward implementing knowledge about aging, metacognition, and strategy use gained in the last funding cycle in a new and unique training program for older adults that simultaneously restructures negative beliefs about age and learning, trains relevant strategies for learning, and trains the use of self-testing (a practical method for using monitoring to guide study) during learning. Area 2 builds on work completed or in progress indicating age-related sparing of the ability to monitor the cognitive system. The investigators focus on factors that influence both relative and absolute accuracy of monitoring, and evaluate age differences in spontaneous (uninstructed) strategy use. Area 3 directs attention to the nature of mediational strategies used during associative learning, extending work with new methods for measuring strategy self-reports developed in the last funded cycle to evaluate age differences in spontaneous (uninstructed) strategy use. Area 4 evaluates the utilization of monitoring to guide self-paced study of paired-associates, in part through the creation of new metacognitive judgements that should help to explain age differences already identified in the utilization of monitoring. Outcomes obtained from all four areas have important theoretical implications for age differences in associative learning and will also provide valuable information on how to construct training and interventions programs to help older adults optimize learning in everyday situations.